346 Prof. \V. N. Hartley. On Var'u <<;<> /,*., /,-,/ [Apr. 1! 



The origin of these coincident portions of spectra, namely, from th 

 combnstion of cyanogen and from carbon electrodes in saline solu< 

 t ii -us, taken in conjunction with the fact that they arc not rende: 

 by cyanides, makes it doubtful whether the cyanogen spectrum 

 not due to elementary carbon, as first advocated by Marshall War 

 There are other facts and circumstances which somewhat snppo 

 this doubt. First, variations have been observed in the spectrum o: 

 carbon which cannot be easily accounted for. Secondly, the effect 

 one substance on the spectrum of another which I have recentl 

 observed not only strengthens weak lines but in certain cases bri 

 a new series of lines into view. Thirdly, the spectra of mix 

 vapours have been shown to be different from the spectra of the su 

 stances by themselves (Liveing and Dewar, 'Boy. Soc. Proc.,' 

 34, p. 428) ; and, fourthly, the influence of the strong lines of 

 element on adjacent weaker lines of another substance is to strengthei 

 the weaker lines in some cases, but almost to obliterate them 

 others. 



In order the more readily to be able to refer to modifications in 

 the carbon spectrum, I append lists of the lines which appear in 

 different circumstances when condensed sparks are passed between 

 graphite electrodes, so that at a glance it may be noted in whal 

 manner the spectra are modified. For comparison the edges of the 

 cyanogen bands measured by Kayser and Runge, and the carbon 

 lines as recorded by Eder and Valenta, are given. The differem 

 between the wave-lengths quoted for the same lines is due chiefly t 

 the different scales used, namely those of Angstrom and of Rowlan 



On examining these figures it will be observed that Hartley a 

 Adeney's carbon line, 3881'9, may or may not coincide with the edj 

 of the cyanogen band, 3883'6, of Kayser and Runge, but probabl; 

 it does not. The carbon line, 3870'7 appears to coincide with 3871*7, 

 the band of cyanogen, but there is no further agreement between 

 two spectra until we come to the three lines of carbon 3589*9, 3584*8, 

 and 3583-5. The two lines 3881'9 and 3870'7 of carbon are obtain* 

 when dry electrodes are immersed in carbon dioxide, and under this 

 condition the second line is greatly lengthened and both are much 

 strengthened, whereas in atmospheric air, though the first is long, 

 both are only feeble. With wet electrodes in air they are both long. 

 With dry electrodes in oxygen 3881*9 is a long line, but 3870'7 it 

 absent. 



I conclude from these facts that, whatever may be the origin 

 these lines, they do not arise from the presence of any compo 

 of carbon with nitrogen, while at the same time they do appear 

 belong to the element carbon. 



Of the three lines given above, that with the wave-length 3589 

 appears when sparks are passed between dry electrodes of graphite 



